One of the fundamental questions of modern immunology is the control of CD4+ helper T-cell differentiation into subsets with discrete effector functions. Of special interest is how the subsets interact with each other to promote strong immunity and lifelong stable immunological tolerance. A variety of investigations show that there are at this time at least five well-defined subsets: Th0, Th1, Th2, Treg, and Th17 that can both promote and inhibit the behavior of each other. Recently, a great deal of experimental attention has been directed at the subset of CD4+,CD25+ T cells that express high levels of the winged fork head transcription factor, FoxP3. This subset strictly limits their own expression of proliferative cytokines and is generally incapable of producing growth cytokines such as IL-2 and IL-4. These cells have been shown to impair the activation of other conventional CD4+ T cells which could be important during autoimmune reactions; hence these cells are called T regulatory (Treg) cells. Despite a great deal of investigation, it is still unclear by what molecular mechanisms Treg cells are capable of suppressing other subsets of T cells and what role they play in normal immune reactions. In order to examine the molecular basis of the suppressive effects of Tregs, we investigated a co-cultivation system in which Tregs were used to suppress conventional CD4 T cells responding to T cell receptor agonists. We found that potent suppression could be observed with 1:1 mix of cells. Under these conditions, we observed that the conventional T cells underwent apoptosis. Furthermore, we found that the prevailing model of suppression of IL-2 gene transcription was incorrect. Death of the responding cells could quantitatively account for the loss of T cell response. We found that various common gamma chain cytokines were able to completely reverse the death due to Treg suppression. This appeared to be due to the fact that Tregs consumed, but could not produce, the cytokines and thereby deprived the conventional T cells of cytokines. In conjunction with these experiments, we found that a deficiency of the Bim gene completely rescued T cells from Treg suppression. We also demonstrated that similar apoptosis and T cell deletion effects could be observed in vivo in an inflammatory bowel disease model. Taken together, these data suggest that Treg cells exert their suppressive effect by a form of polyclonal deletion rather than suppression of cytokine transcription in the responding conventional T cells. [unreadable] We also discovered that Treg cells express receptors for gamma chain cytokines and are dependent on an exogenous supply of these cytokines to overcome cytokine withdrawal apoptosis in vitro. This result was validated in vivo by the accumulation of Treg cells in Bim-/- and Bcl-2 tg mice which have arrested cytokine deprivation apoptosis. We also found that CD25 and Foxp3 expression were down-regulated in the absence of these cytokines. CD25+ cells from Scurfy mice do not depend on cytokines for survival demonstrating that Foxp3 increases their dependence on cytokines by suppressing cytokine production in Treg cells. Our study reveals that the survival of Treg cells is strictly dependent on cytokines and cytokine producing cells because they do not produce cytokines. Our observations shed new light on this subset of T cells and we are now attempting to develop a general model of how T cell subsets achieve tolerance and immunity. These investigations may have importance for autoimmune disorders, infectious diseases, and tumor immunity.